Mike Chaney's Tech Corner
December 28, 2024, 06:02:32 PM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News: Qimage registration expired? New lifetime licenses are only $59.99!
 
   Home   Help Login Register  
Pages: [1] 2
  Print  
Author Topic: Profiling my Canon scanner  (Read 33144 times)
JohnF
Newbie
*
Posts: 22


Email
« on: August 15, 2009, 04:01:13 PM »

Hello.
I just received my Profileprism software and am trying to profile my Canon 5000F scanner.
I have Vuescan set to the settings recommended in the ProfilePrism help file and am outputting a raw .tif file.
PP is always telling me that the colour balance is incorrect by 7-8% and some shadow areas are clipped.
However, changing the WB/colour/brightness settings in Vuescan does not improve matters and I am going around in circles.

Also, altering the brightness/contrast/colour settings in PP do not change the histogram.
What, if anything, am I doing wrong?
Thanks in advance.
John Huh?
Logged
Terry-M
The Honourable Metric Mann
Forum Superhero
*****
Posts: 3251



WWW
« Reply #1 on: August 15, 2009, 07:47:39 PM »

Quote
PP is always telling me that the colour balance is incorrect by 7-8% and some shadow areas are clipped.
However, changing the WB/colour/brightness settings in Vuescan does not improve matters and I am going around in circles.
Hi John,
I profiled my Canon 8400F some while ago and seemed to get reasonable results. I've just looked at the results again and included a screen snap as an attachment below.

I too used VueScan, did you output a 48bit raw tiff file? You should not make any adjustments in Vuescan, just use the raw, unadjusted scan. Also, did you calibrate the scanner from Vuescan "Scanner"  first?

I note from looking at my results again, I set "Allow Profile to Fix WB"

In general, your results sound ok. How many patches had an X?
Terry.
Logged
JohnF
Newbie
*
Posts: 22


Email
« Reply #2 on: August 16, 2009, 09:45:22 AM »

Thank you Terry for that reply.
My Vuescan settings are as follows.
Scanner - Calibrate
Input - Bits per pixel - 48 bit RGB
Input - Scan resolution - 300dpi
Color - Color balance - None
Color - Scanner color space - Built-in
Color - Output color space - sRGB
Output - Printed size - Scan size
Output - Only raw file ticked
Output - Raw file type - 48 bit RGB
Output - raw output with - Scan

Those are the only ones I think are relevant to this project unless you know differently Wink
This is what I got with these settings.

I assume that before I profile my printer it is essential to get the scanner profile correct?
Many thanks again.
Regards
John

Logged
JohnF
Newbie
*
Posts: 22


Email
« Reply #3 on: August 16, 2009, 10:04:18 AM »

Since posting the above I have done some more tests in PP.  I set Brightness to -8, Contrast to -8 and Saturation to +8 and the result was exactly the same.  That cannot be right surely?  Something odd going on don't you think?
Regards
John
Logged
JohnF
Newbie
*
Posts: 22


Email
« Reply #4 on: August 16, 2009, 11:14:22 AM »

Some more tests.
Using my former Vuescan settings I created a new scan and imported it into PP using the Camera/Scanner-Prism target setting.
The result was:
Exposure image is properly exposed (247)
WB accuracy 2%
Lighting variance 5%
Warning 14 patches marked with X
Shadow detail - Poor 3 or more dark grey patches are clipped.

I changed PP setting to Printer-Prism target and imported the same file.
The result was:
Exposure image is properly exposed (252)
WB accuracy 2%
Lighting variance 5%
Warning 4 patches marked with X
Shadow detail - Excellent- no grey patches clipped.

I found this out earlier by going ahead and creating a Printer profile which produced excellent results with an earlier scan.
I am confused as to why changing the PP setting from Camera/scanner to Printer should give such different results with the same file.
Regards
John
Logged
Seth
Sr. Member
****
Posts: 322



« Reply #5 on: August 16, 2009, 02:57:41 PM »

What chart and IT8 file did you use when you first profiled Vuescan?

I would think you'd want to get Vuescan back to default if you are going to use PP as the scanner profiler.  Surely it comes with it's own chart and IT8 file. 

After using PP, you'll have to choose ICC profile and set the the new box that opens to the profile you've made.  You'll also have to point Vuescan to the IT8 file for the chart you've used instead of its internal file.
Logged

Seth
<CWO4 (FMF) USN, Ret.>
JohnF
Newbie
*
Posts: 22


Email
« Reply #6 on: August 16, 2009, 04:12:59 PM »

What chart and IT8 file did you use when you first profiled Vuescan?
I used the chart that came with ProfilePrism and put the Vuescan settings to those recommended in the help files.
Because the resultant scan was so poor (according to PP) I never used it.
Thanks for your input
John
Logged
Terry-M
The Honourable Metric Mann
Forum Superhero
*****
Posts: 3251



WWW
« Reply #7 on: August 16, 2009, 08:46:04 PM »

Quote
My Vuescan settings are as follows
They look ok to me.

Quote
I assume that before I profile my printer it is essential to get the scanner profile correct?
I don't think so (I haven't), because you use a raw tif file for printer profiling, colour space conversion does not come into the process.
Mike would have to confirm this etc.

Quote
I used the chart that came with ProfilePrism and put the Vuescan settings to those recommended in the help files.
Because the resultant scan was so poor (according to PP) I never used it.
Not sure what you mean by this, you should use the target supplied and specify that in the PP dialogue.
I do notice from your screen shot, the Reference target specified is not the same as mine. The reference is on the bottom right of the IT8 target and what you select in PP must match that reference. Is that your problem?

Quote
I am confused as to why changing the PP setting from Camera/scanner to Printer should give such different results with the same file.
I wonder if it's related to the IT8 target reference not being correct when you made a scanner profile.

I note from your PP output data, the histogram is heavily biased to the left, ie. under exposed, so you will get shadow detail problems.
However all this is irrelevant if you specified the wrong reference in PP.
Please check that and report back.


Terry.
Logged
Seth
Sr. Member
****
Posts: 322



« Reply #8 on: August 16, 2009, 09:23:13 PM »

Terry, et al-

I use Xrite/GM Eye1 Pro, so am not familiar with the PP systems.  However, I do not understand doing the profiling THROUGH Vuescan.  Doesn't PP have its own profiling system that would scan while bypassing Vuescan?

It just seems there are two iterations going on here-- PP and the Vuescan IT8 file.

For example, when printer profiling, iMatch gives the option to print the chart through it or Photoshop or QI--my choice.  But, iMatch does not print THROUGH one of the others.  I hope I am making myself clear.

My take from what I am hearing is the scanner profile should be done directly wit PP.  THEN go into Vuescan and enter the newly created profile in the scanner profile box.  And replace the Vuescan IT8 file with the one that comes with the PP chart.  Make sense?
Logged

Seth
<CWO4 (FMF) USN, Ret.>
Terry-M
The Honourable Metric Mann
Forum Superhero
*****
Posts: 3251



WWW
« Reply #9 on: August 16, 2009, 09:37:50 PM »

Quote
so am not familiar with the PP systems
Right. Here is quick summary of what PP does with respect to a scanner profile and the involvement of Vuescan.

PP is a scanner based profiling software uses an IT8 target. Vuescan is used as the scanning software to make an image of the target for PP to make the profile. PP contains it's own reference files for various targets.
Because Canon scanner software is not very good for producing target scans, Vuescan is recommended,. That's the relevance of Vuescan in this situation although it can be used to produce profiles too, but to go into that will just confuse matters here.
Terry.
Logged
JohnF
Newbie
*
Posts: 22


Email
« Reply #10 on: August 17, 2009, 10:48:48 AM »

Thanks guys for your input.  Terry.  I am using the correct IT8 chart supplied with my software.  It's a different number to yours.

I was firstly trying to profile the scanner, thinking I would HAVE to do that before making the printer profile but apparently not. I scanned the chart supplied as a RAW .tiff with all the settings recommended in the help file but was not sure as to what colour space I should use for the scanner so used 'built-in'.  In Vuescan's Color tab I selected NONE for the Color Balance. I was never able to get a really satisfactory result from PP as detailed above so went on to create a printer profile.

Here is where I am puzzled because using the same Vuescan settings as for scanner profiling, selecting Printer profiling in PP, gave me an excellent result with just 3 xs. The resultant profile produced excellent results from the test images printed from Qimage, easily the best I have ever seen them so much progress has been made there.

So, I suppose there is really no need to profile my scanner but I would like to get to the bottom of why I could not get a good result from the scanning profile project.

Further thoughts/comments greatly appreciated.
Regards
John

Logged
Terry-M
The Honourable Metric Mann
Forum Superhero
*****
Posts: 3251



WWW
« Reply #11 on: August 17, 2009, 10:55:37 AM »

John,
I think we need MIke's input now.
terry.
Logged
Fred A
Forum Superhero
*****
Posts: 5644



WWW Email
« Reply #12 on: August 17, 2009, 11:05:20 AM »

Sorry to butt in, but you do not use the scanner profile when scanning for a printer profile.
The scanner profile is used when you are scanning anything other than targets.
Scanning a target, no scanner profile.

So essentially, you just want to make a scanner profile, just because......
That's fine... there is a secret ingredient embedded in there that is unpredictable. The scanner itself.
Some have lighting that exceeds the ability of the software to cope.
Sounds like you have a winner there.... printer profiles are the toughest, and you made good ones.
Fred
Logged
Fred A
Forum Superhero
*****
Posts: 5644



WWW Email
« Reply #13 on: August 17, 2009, 11:06:53 AM »

John,
I think we need Mike's input now.
terry.

Why should Terry get an extra post because he posted the same message twice.
Hmmmm. No wonder his posting totals are so high.  Tongue Tongue Tongue Tongue Tongue
Logged
Terry-M
The Honourable Metric Mann
Forum Superhero
*****
Posts: 3251



WWW
« Reply #14 on: August 17, 2009, 11:17:18 AM »

Quote
Why should Terry get an extra post because he posted the same message twice.
No worries, I've already got the "Senior" badge  Grin

Quote
Some have lighting that exceeds the ability of the software to cope.
Now we have confirmed the profile reference is correct, that would have been my next comment. The histogram was very much biased to the left.
John, I assume your printer profile histograms look good? Have you printed some test images yet with the new PP printer profile?
Terry
Logged
Pages: [1] 2
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Valid XHTML 1.0! Valid CSS!
Security updates 2022 by ddisoftware, Inc.